The neural circuitry involved in processing the visual stimulus of a human face has received considerable research attention, with evidence indicating that some brain regions may respond preferentially to such stimuli. However, we know little about how neural processing is affected by the subject's familiarity with, or emotional attachment to, the person whose face is depicted. Research in this area could form a bridge between cognitive neuroscience and the newly emerging field of affective neuroscience by indicating pathway(s) linking limbic and sensory circuits. In addition, the use of emotionally significant faces as activation stimuli may help to elucidate the neural circuitry underlying interpersonal attachment and psychological responses to loss. Thus, this functional magnetic resonance imaging protocol uses a face recognition activation task to examine how neural processing varies depending on whether the face depicted is emotionally significant to the viewer. In a series of studies, subjects view unfamiliar and familiar (including personally familiar and famous) faces; mothers view pictures of their own child and those of familiar and unfamiliar children; and bereaved and non-bereaved spouses view pictures of their spouse, other family members, and unfamiliar people. To date, ten subjects have completed the study that uses personally familiar, famous familiar, and unfamiliar faces as stimuli, and ten mothers have completed the second study. Data from the first study indicate that amygdala activation is decreased, and anterior superios frontal and fusiform activation increased, when subjects view personally familiar faces. In addition, seven spousally bereaved individuals and five matched controls have completed the third study. Recruitment of bereaved subjects and controls is continuing.